Galactic Bulge Time Domain Survey (GBTDS)

• The Galactic Bulge Time Domain Survey (GBTDS) is a high-cadence, multi-epoch imaging program that monitors the inner Milky Way with precise photometric and astrometric data.
• It employs space- and ground-based observations, using microlensing, transits, and asteroseismology to discover exoplanets, variable stars, and compact binaries.
• Synergies with LSST and Euclid mitigate systematics and enhance studies of stellar populations, Galactic structure, and gravitational-wave sources.

The Galactic Bulge Time Domain Survey (GBTDS) is a set of planned and proposed observational programs, primarily associated with the Nancy Grace Roman Space Telescope and the Rubin Observatory's LSST, that will perform high-cadence, high-precision, wide-area photometric and astrometric monitoring of the inner Milky Way. The primary objectives include exoplanet discovery and characterization via microlensing and transits, variable star and compact binary census, asteroseismology, resolved Galactic structure, and gravitational-wave science in the microhertz regime. GBTDS delivers multi-epoch, near-infrared and optical time-series imaging, enabling unique probes of stellar populations, exoplanet demographics, nuclear star-cluster dynamics, and multi-messenger astrophysics.

1. Survey Architecture and Observing Parameters

GBTDS, in its reference implementation on the Roman Space Telescope, monitors the central $\sim$2 deg$^2$ of the Galactic bulge using the Wide Field Imager (WFI), covering seven contiguous fields ($0.281\,\mathrm{deg}^2$ per field) with 15-min cadence in the broad F146 (0.9–2$\ \mu$m) filter, interleaved with other bands for color characterization (Yee et al., 2023, Terry et al., 2023). The nominal mission schedule includes six observing seasons, each 60–72 days, over five years for a cumulative on-sky time of 360–432 days. The survey reaches $\gtrsim 100$ million stars down to F146$\approx$21–24 mag, with $\sim$1–2 mmag precision for bright sources and $\sim$0.1 mag at the faint end (Digman et al., 2022, Pardo et al., 2023).

In parallel, LSST proposes a $\sim$600–760 deg$^2$ optical-only time-domain footprint ($-15^\circ < \ell < +15^\circ$, $-10^\circ < b < +10^\circ$), using $i$-band for cadence ($\sim$18 visits/yr/field) and $g,r,i,z,y$ bands for deep color-magnitude mapping (Gonzalez et al., 2018, Bono et al., 2018). Dual “shallow” and “deep” LSST sub-surveys complement Roman with high-cadence optical variability and proper-motion mapping.

2. Microlensing Exoplanets and Planetary Discovery Space

GBTDS is designed to conduct a statistical census of cold, low-mass exoplanets through gravitational microlensing in the bulge, extending planetary population measurements beyond the snow line, down to or below Earth mass, and including the potential for free-floating planet detection (Yee et al., 2023, Weiss et al., 6 Mar 2025, Terry et al., 15 Oct 2025). The program is optimized for three critical discovery spaces:

• Wide-orbit planets: Roman will uniquely enable the detection of planets at projected separations $s > 2.5\,\theta_E$ (log $s > 0.4$), well into the analogs of the outer Solar System (e.g., Uranus/Neptune), by leveraging high S/N and low blending at $A\sim1$ events (Yee et al., 2023).
• Small mass-ratio planets: High-cadence, low-noise light curves make the survey sensitive to planets with log $q < -4.5$, corresponding to Mars-mass or smaller planets around typical bulge stars.
• Free-floating planet candidates: Short-timescale, isolated events with directly measurable finite-source effects ($\rho \gtrsim 1$) enable constraints on the lowest-mass population.

The expected event rate is $\sim$3,300 microlensing events per field per year, for a total of $\sim$27,000 over six seasons in the baseline survey (Kerins et al., 2023). The yield for bound exoplanets is $\sim$1,400, with hundreds of free-floating planet candidates if such a population exists at the $1\,\mathrm{deg}^{-2}$ level (Kerins et al., 2023, Terry et al., 15 Oct 2025). Approximately 40% of detected planet hosts will have mass and distance measurements with $<20\%$ precision per the formal science requirements (Terry et al., 15 Oct 2025), with measurement pipelines validated against HST/Keck legacy light curves.

3. Stellar and Asteroseismic Population Studies

GBTDS provides unprecedented photometric time-series for millions of bulge/disk stars, enabling variable/star tracer studies and asteroseismic analysis (Weiss et al., 6 Mar 2025, Kerins et al., 2023, Gonzalez et al., 2018, Bono et al., 2018). Simulations show that, for a 15-min cadence, the detection probability for $\nu_\mathrm{max}$ in red giant and clump stars is $>80\%$ at $F146<16$ mag, yielding $185,000$–$349,000$ asteroseismic detections in the bulge under realistic noise and cadence combinations (Weiss et al., 6 Mar 2025). The survey will enable full 3D mapping of the bulge/bar structure via RR Lyrae, Cepheid, Red Clump, and Mira tracers, with population yields summarized in the table below:

Tracer	Expected Yield	Age Range	Primary Use
RR Lyrae	$\sim38,000$	>10 Gyr	Old spheroid, 3D structure
Type II Cepheids	several 1000	>10 Gyr	Old spheroid
Red Clump	millions	1–9 Gyr	Box/peanut bar, distance ladder
Classical Cepheids	500–1000	10–300 Myr	Recent SFH, disk mapping
Miras	thousands	0.5–10 Gyr	Bar/disk overlap, population age

GBTDS is also expected to detect and characterize minute-timescale variability in compact objects, flaring M dwarfs, and short-period pulsators (Kupfer et al., 2023, Digman et al., 2022).

4. Transiting Exoplanet Yields and Demographics

Pixel-level and injection/recovery simulations predict the Roman GBTDS will discover between $\sim$60,000 and $\sim$200,000 transiting exoplanets (Wilson et al., 2023, Kerins et al., 2023). Of these, $7,000$–$12,000$ will be small planets ($R_p<4\,R_\oplus$), with the yield tightly determined by systematics in cadence, duration, and the metallicity distribution function of the underlying stellar populations. For instance, among mid-M and UCDs ($\sim$100,000 such hosts brighter than F146=21), the predicted yield is $1347^{+208}_{-124}$ small transiting planets at $7.1\,\sigma$ or better significance, including $13^{+4}_{-3}$ habitable terrestrial planets ($R_p<1.23 R_\oplus$); discrepancy from these numbers would challenge current occurrence-rate extrapolation, probing planet formation at the bottom of the main sequence (Tamburo et al., 2023).

The “RoSETZ” augmentation would enable a SETI-optimized search of the Earth Transit Zone, predicting 120–630 new habitable-zone, Earth-sized transits around K/M stars—5–20$\times$ the current sample and transformative for $\eta_\oplus$ measurement and demographic modeling (Kerins et al., 2023).

5. Galactic Center, Globular Clusters, and Extended Science

Adding a high-cadence GC field (0.281 deg$^2$ at Sgr A*) increases the survey’s reach to 3.3 million stars in the nuclear star cluster/disk environment (Terry et al., 2023). Yields include $\sim$25,000 microlensing events, $\sim$180 planets (33 Earth-mass, 25 FFPs), $\sim$350 compact binaries (including LISA-band GW sources), and up to 28,000 exoplanet transits. Proper motion and parallax measurements reach precisions of $2.5$–$3.5\,\mu\mathrm{as/yr}$ and $3\,\mu$as, respectively.

Simultaneous monitoring of globular clusters NGC 6522 and NGC 6528 enables a $>3\sigma$ test of planet occurrence as a function of metallicity and cluster environment, predicting $13\pm4$ detectable hot Jupiters for the metal-rich cluster and $\approx1$ for the metal-poor (Grunblatt et al., 2023), with scientific leverage also for variable star and binary fraction studies.

6. Gravitational-Wave and Multi-Messenger Astrophysics

GBTDS is uniquely poised to pioneer gravitational-wave science in the microhertz regime, bridging the gap between PTA and LISA bands via relative astrometry (Pardo et al., 2023). Its $10^8$ monitored stars, $\sim$41,000 exposures/star, and per-epoch astrometric precision of $\sim$1.1 mas yield strain sensitivities $h_c\sim 10^{-15}$ at $10^{-4}\,\mathrm{Hz}$. If the mean field-deflection is recoverable, the survey can detect the stochastic SMBH binary background at SNR$\sim$70 and individual binaries with chirp mass $10^9\,M_\odot$ out to 1 Gpc.

GBTDS will discover $\sim$4–6 detached LISA-band double white dwarf binaries per season, with $\gtrsim$1.5 being Romand+LISA joint “verification” binaries, depending on tidal-heating uncertainties (Digman et al., 2022). Multi-messenger campaigns are enhanced by coordinating with Euclid for proper-motion and parallax improvements, directly increasing planet mass measurement precision by up to $\times 5$ (Kerins et al., 2023).

7. Synergies, Systematics Mitigation, and Survey Optimization

Integration of GBTDS with LSST (optical) and Euclid (space-based NIR) provides mitigation of blend, extinction, and parallax systematics (Kruszyńska et al., 20 Jun 2024, Kerins et al., 2023, Kerins et al., 2023). For example, complementarity in wavelength affords extinction-resilient source flux determination; cross-survey astrometry enables improved proper-motion cleaning, and dual-line-of-sight field-pairing (such as GBTDS with RoSETZ) enables empirical de-blending of transit depths and joint constraints on microlensing model priors.

Trade studies and survey design optimization leverage metrics such as detection completeness as a function of cadence, field coverage, and crowding. For exoplanet, binary, and variable-star metrics, analytic and empirical simulation pipelines evaluate detection efficiency, parameter degeneracy rate, and the impact of re-allocating time between fields or cadences.

---

References:

(Tamburo et al., 2023, Yee et al., 2023, Terry et al., 15 Oct 2025, Weiss et al., 6 Mar 2025, Digman et al., 2022, Wilson et al., 2023, Kerins et al., 2023, Grunblatt et al., 2023, Bono et al., 2018, Gonzalez et al., 2018, Terry et al., 2023, Pardo et al., 2023, Kerins et al., 2023, Kruszyńska et al., 20 Jun 2024, Kupfer et al., 2023)